tensorflow / tensorflow/probability
A bug in Linear_Mixed_Effects_Models.ipynb
Open
Nobody has claimed this yet.
- Dominant language
- Jupyter Notebook
- Stars
- 4.4k
- Forks
- 1.1k
- PR merge metrics
- No merged PRs in 30d
Description
There seems to be a bug in the model specification. in
tfd.MultivariateNormalDiag(
loc=tf.zeros(num_students),
scale_diag=self._stddev_students * tf.ones(num_students)),
tfd.MultivariateNormalDiag(
loc=tf.zeros(num_instructors),
scale_diag=self._stddev_instructors * tf.ones(num_instructors)),
tfd.MultivariateNormalDiag(
loc=tf.zeros(num_departments),
scale_diag=self._stddev_departments * tf.ones(num_departments)),
it seems that self._stddev_students, self._stddev_instructors, and self._stddev_students are not being tracked by the GradientTape and therefore not updated properly in the m step.
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with Linear_Mixed_Effects_Models.ipynb and inspect the model specification around the three MultivariateNormalDiag distributions. Trace how GradientTape observes self._stddev_students, self._stddev_instructors, and self._stddev_departments during the M step; the fix is complete when these parameters are tracked and updated correctly.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- jupyter-notebook, python
- Domain
- machine-learning
- Issue type
- Bug
- Difficulty
- 3/5
- Estimated time
- 1-2 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100